A, B - unistdatamining

advertisement
Rediscovering Workflow Models from
Event-Based Data using Little Thumb
Han-na Yang
Introduction

Work flow management systems


Offer generic modeling and enactment capabilities for structured business
processes

Too restrictive

Have problems dealing with change (=not flexible)

ex) Staffware, IBM MQSeries, COSA, etc.
Many problems are resulting from discrepancy between workflow design
and workflow enactment
“Reverse the process”: Process Mining

We start by gathering information about the workflow processes as they
take place. Do not start with a workflow design.



Assumption: it is possible to construct workflow logs with event data
Process mining


Any information system using transactional system or workflow management
system will offer this information in some form
The method of distilling a structured process description from a set of real
executions.
We focus on workflow processes with concurrent behavior

Detecting concurrency is one of our prime concerns

Distinguishing AND/OR splits/joins explicitly
Workflow Process Model


Workflows

‘Case-based’: every piece of work is executed for a specific case

Case: handled by executing tasks in a specific order (ex. Insurance claim, mortgage)
Workflow process model

Specifies which tasks need to be executed and in what order

Routing elements: describe sequential, conditional, parallel and iterative routing
Petri nets

Tasks are modeled by transitions

Places and arcs model causal dependencies

Split & Join
OR-join
OR-split
AND-split
AND-join
WorkFlow net (WF-net)




A Petri net that models the control-flow dimension of a workflow
Focuses on the process perspective and abstract from the functional,
organization, information and operation perspectives
Sound WF-nets

Termination is guaranteed

No dangling tokens are left behind

No dead task
Workflow log is sequence of events
A Heuristic Process Mining Technique


Four ordering relation in the α-algorithm
(Let A, B be events, W a workflow log)

A>B: if and only if there is trace line in W in which event A is directly followed by B

A→B: dependency relation, B depends on A

A#B: non-parallel relation, no dependency between A and B

A∥B: parallel relation, used to detect the kinds of splits and joins
Heuristic mining technique

Less sensitive for noise and the incompleteness of logs than α-algorithm

Three mining steps
①
The construction of a dependency/frequency table(D/F-table)
②
The induction of a D/F-graph out of a D/F-table
③
The reconstruction of the WF-net out of the D/F-table and the D/F-graph
Step 1: Construction of the dependency/frequency table

For each task A these information is abstracted out of the workflow log
①
The overall frequency of task (#A)
②
The frequency of task A directly preceded by another task B (#B>A)
③
The frequency of A directly followed by another task B (#A>B)
④
⑤

A local metric that indicates the strength of the dependency relation between task
A and another task B ($A→LB)
A more global metric that indicates the strength of the dependency relation ($A→B)
$A→B-dependency counter

It is incremented with a factor (δ)n

Dependency fall factor(δ: delta) is [0.0 … 1.0]

n is the number of intermediary events between them

Therefore, if task B appears directly after task A then (δ)n=1(n=0).

$A→B-dependency counter decreases if the distance between tasks increases.
Step 1: Construction of the dependency/frequency table
>

Example

T6 is never directly
preceded by T10
(#B>A=0)

T6 is often directly
followed by T10
(#A>B=581)
Step 2: Induction of dependency/frequency graphs

Heuristic rules
>




Four conditions demand that specific values of the D/F graph (#A>B, #B<A, $A→LB,
$A→B) are higher or lower than a certain threshold value(σ, N1, N2)
Only task-pattern occurrences above a threshold frequency are reliable enough for
our induction process
Formulating a rule that for each pair of events A and B takes the decision if they
are in the dependency relation or not is not really necessary.
First (temporally) version of mining rule 1. given a task A:


A→ X if and only if X is the event for which DS(A,X) is maximal
Y→A if and only if Y is the event for which DS(Y,A) is maximal

Dependency score: DS(X,Y) = (($X→LY )2 + ($X → Y )2) / 2

New rule does not contain any parameter
Step 2: Induction of dependency/frequency graphs

For each arc the dependency score(DS) is given and for each task the
number of event occurrences in the log
Step 2: Induction of dependency/frequency graphs

The first version of mining rule 1 is updated


Mining rule 1 (definite version). Given a task A
 Suppose X is the event for which DS(A,X)=M is maximal. Then A→Y if and only if
DS(A,Y)<0.95*M
 Suppose X is the event for which DS(X,A)=M is maximal. Then Y→A if and only if
DS(Y,A)<0.95*M
Threshold value is(0.95) only one parameter and the parameter seems
robust for noise and concurrent processes.
Step 3: Generating WF-nets from D/F-graphs

The types of the splits and joins are not represented in the D/F-graph

Useful information to indicate the type of a join and split


Information in the D/F-table: determine split or join

A to B AND C (AND-split): pattern B,C and pattern C,B can both appear

A to B OR C (OR-split): pattern B,C and pattern C,B will not appear
The frequency of the nodes in the D/F-graph


Used for the validation of the induced workflow model
After observations, apply the α-algorithm to translate this information
into a WF-net
Little Thumb

A tool that attempts to induce a workflow model from a workflow log



The workflow log may contain errors(noise) and can be incomplete
Steps to analyze the loaded workflow log
①
Already executed (D/F-table in the Figure 4)
②
Induction of a D/F-graph out of D/F-table (Figure 5)
③
Use the information in the extended D/F-table to indicate join and split (Figure 6)
Check WF-net tab: possibility to validate the WF-net


First check: checks if the trace can be parsed by the WF-net
Second check: test out the frequency information of the events is in accordance
with the structure of the minded WF-net
Little Thumb
Generate WF-log tab


It is possible to load a
WF-net and to generate
workflow logs, with or
without noise
Select-events tab


We can concentrate our
mining process on the
most frequent events
and neglect low
frequent events
Little Thumb

The types of the splits and joins are not represented in the D/F-graph
Little Thumb
First Experiments

Six different free-choice workflow models

All models contain concurrent processes and loops


For each model we generate three random workflow logs with
1000event sequences

A workflow log without noise

One with 5% noise

A log with 10% noise
Four different types of noise

Delete the head of a event sequence

Delete the tail of a event sequence

Delete a part of the body

Interchange two random chosen events
First Experiments
>


Six noise free workflow logs results in six perfect D/F-graphs
If we add 5% or 10% noise to the work flow logs, the resulting D/Fgraphs and WF-nets are still perfect
Second experiments

Four elements strongly influence the behavior of a WF-net and/or the
workflow mining process

The number of event types in the WF-net


The amount of material in the workflow log



100, 200, 600, 1000, 1400, 2000 trace lines
The amount of noise


12, 22, 32, 42 event types
5%, 10%, 20%, 50% noise
The unbalance res to the probability that enabled event will fire
Generate 480 different workflow logs by varying each of the above
enumerated elements
Second experiments

Conclusion


Under all circumstances most dependency relations, the type of split and the type
of joins are correctly found
Mining technique appears especially robust for the number of trace lines and the
amount of unbalance

50% noise cause serious problems

Most errors have to do with short loops

An improvement of the heuristic rules for short loops seems necessary
ProM Tool (1)
ProM Tool (2)
ProM Tool (3)
ProM Tool (4)
Questions ?
Download